Most cancer deaths are caused by metastasis, or spread, of the cancer to distant organs. By the time that metastasis has advanced to the point where it can be detected, the number of cancer cells in the body has doubled as many as 30 times, resulting in a large tumor burden of genetically deregulated cells that are highly resistant to treatment. Recent work has shown that molecular information or biomarkers obtained from the primary tumor at the time of initial diagnosis can identify which patients are likely to harbor undetectable micrometastasis. This would allow adjuvant, or preemptive, treatment of these high-risk patients rather than waiting for the cancer to double many times and become more resistant to therapy. Ocular (uveal) melanoma is one such cancer where a highly accurate biomarker of micrometastasis is needed. Up to 50% of ocular melanoma patients develop metastatic disease that is invariably fatal despite successful treatment of the primary eye tumor, indicating that micrometastasis was already present in those patients when the primary eye cancer was treated. Metastatic disease usually becomes detectable 2-5 years later, which provides a window of opportunity for adjuvant systemic therapy in high risk patients. A highly predictive biomarker for ocular melanoma derived from the primary tumor was developed during the previous funding period. This biomarker is an assay that measures the coordinate expression of 12 marker genes and 3 control genes, a so-called gene expression profile. Computer programs that recognize patterns in the profile from individual tumors can distinguish with great accuracy which patients are likely to harbor micrometastasis. This assay is the most accurate predictor of metastasis to date for this cancer and can be used to guide adjuvant therapy in high risk patients. In the grant renewal, the assay will be refined and optimized for routine clinical use by collecting and analyzing a large number of tumor samples prospectively through a multi-center study. Further, the assay will be evaluated and optimized for its prognostic value in cutaneous melanoma by analyzing a large set of archival tumor specimens with long clinical follow-up. The results of this research are likely to improve patient care in ocular melanoma and perhaps in cutaneous melanoma, which is a much more common cancer. More generally, these studies provide a proof of principle for the current major emphasis of the NIH on a personalized, predictive, preemptive approach to the management of complex chronic diseases such as cancer.